Reaction of Catechol with Colloidal Silica and ... - ACS Publications

Apr 12, 1971 - DENNIS W. BARNUM is present as NO-, and NO- is a strong a-donor ligand which competes favorably with NH3 for u electrons. This cobalt ...
0 downloads 0 Views 880KB Size
1424 Inorganic Chemistry, Vol. 11,

6, 1972

is present as NO-, and NO- is a strong a-donor ligand which competes favorably with NH3 for u electrons. This cobalt case is analogous to the situation in [Cr( C T \ ; ) E N O ] ~where - , ~ ~ the axial cyanide is more weakly bound than the equatorial cyanide. In the cobalt example, the two ligands are both competing for u electrons, and NO- is stronger than “3, while in the chromium case the two ligands are competing for H electrons, and NO+ is apparently stronger than CN-. In the present ruthenium case, however, the 8-acceptor NO + ligand does not compete with the u donor C1- ligand, and so both axial bonds (Ru-N and Ru-C1) are strengthened. This conclusion is supported by the structural data reported33 for [Ru(NO)(OH) (N02)4I2+. Using Pauli n g ’value ~ ~ ~of 0.99 A for the covalent radius of chlorine and our average equatorial Ru-CI bond length of 2.376 A we deduce an effective octahedral covalent radius for Ru(I1) in these systems of 1 39 A, which is in fair agreement with the value of 1.33 reported by Pauling 43 A similar calculation, using the average Ru-K02 bond distance33in [Ru(NO)(OH) (N02)4lZ- of 2 079 fi and a covalent radius43for nitrogen of 0.70 A, leads to a value of 1.38 A. Using either of these two calculated radii and a covalent radius for oxygen of 0.66 one predicts that the Ru-0 bond length should be approxi-

a

(42) J. H Enemark, M. S. Quinby, L. L. Reed, M. J. Steuck, and K K Walthers, Inovg Chem , 9, 2397 (1970). (43) L. Pauling, “The Nature of the Chemical Bond,” 3rd ed, Cornel1 University Press, Ithaca, N. Y , 1960.

DENNISW. BARNUM mately 2.04 8. Hence, the axial Ru-0 bond length33 of 1.950 fi in [ R u ( N O ) ( O H ) ( X O ~ ) ~is] ~much shorter than the calculated value; this is again a consequence of the successful competition of the u-donor hydroxide ligand for u electrons, in an exactly analogous manner to that found here for the axial Ru-Cl bond in [Ru(NO)CIE]~--.We would anticipate similar results in other octahedral complexes in which one axial ligand is a T acceptor and the other is a a donor. The structural data are evidently in contradiction to the spectral data, which have been interpreted14 as indicating that the axial Ru-C1 bond is weaker (and, presumably, longer) than the equatorial bonds. We assume that a reexamination of the spectral data, especially if Raman data become available to supplement the infrared data, will lead to the conclusion that the axial Ru-C1 bond is indeed stronger than the equatorial bonds. Acknowledgments.-We are grateful to Professor E. E. Mercer for useful discussions concerning the spectra of this complex. This research was supported by the Advanced Research Projects Agency through Contract DAHC 15-67-C-0223 with the Materials Research Center of the University of North Carolina and by the National Science Foundation through Grant GP-10300; we gratefully acknowledge this financial support. J. T. V. thanks the Tennessee Eastman Corp. for a predoctoral fellowship.

CONTRIBUTION FROM THE DEPARTMEKT OF CHEMISTRY, PORTLAND STATEUNIVERSITY, PORTLAKD, OREGON 97207

Reaction of Catechol with Colloidal Silica and Silicic Acid in Aqueous Ammonia BY DEXNIS W. BARXUM Received April 12, 1971 Reaction of catechol with silicic acid or colloidal silica in 8-15 M aqueous ammonium hydroxide gives large colorless crystals The ammonia can be removed a t room temperature giving a product that analyzes for (NH4)2[Si(C~H.i02)3],H20.This compound has been studied by ir, nmr, X-ray powder diffraction, and its chemical behavior. The “water of crystallization” is probably present as an Si-OH group and a phenolic hydrogen. Removal of this “water” is difficult but can be accomplished in some cases. Other complexes have been prepared containing acetone, ethanol, dimethyl sulfoxide, or a fourth mole of catechol. A complex previously reported by other workers to contain 2 g-atoms of silicon/7 mol of catechol is actually the ethanol solvate (NH4)2[Si(CeHaO2)J.xH*O.CzHsOH where x = 3 or 4.

of a silicon-catechol complex that contains ammonia of crystallization.

Introduction Silicon forms very few coordination complexes with organic molecules and those that are known are generally unstable in the presence of water. An outstanding exception is catechol and other aromatic vic-diols which in aqueous basic solutions react with silica gel, colloidal silica, and even finely divided quartz to form coordination complexes of silicon.l a 2 Although these complexes have been investigated previously, 3--7 modern spectro(1) A. Rosenheim, B. Raibmann, and G. Schendel, Z . Anovg. Chem., 196, 160 (1931). (2) A. Weiss, G. Reiff, and A. Weiss, Z . AmYg. Allg. Chem., 311, 151 (1961). (3) A. Rosenheim and 0. Sorge, Bev., 63,932 (1920). (4) H. Bartels and H. Erlenmeyer, Helw. Chim. Acta, 4 7 , 7 (1964). (5) H. Baumann, Beitv. Silikose-Fovsch,, Soitdevb., 4 , 43 (1960). (6) R. Hess, R . Bach, and H. Deuel, Expevientia, 16, 38 (1960) (7) J. J. Flynn and F. Boer, J . Amel’. Cheni. Soc., 91, 5756 (1969).

metric methods have not been applied. In this investigation difficulty was encountered in repeating reported preparations and much of the previous work has been found to require modification. Rosenheim, et aZ,,I first reported the reaction of an aqueous solution of catechol with freshly precipitated silica in the presence of a base. They prepared [Si(cat)s].9H20* and several other salts of the Si(cat)?-

H2cat

Heat-

cat *-

(8) I t will be convenient in this paper to use the abbreviations Hlcat, Hcat-, and cats- for catechol and its ions

Inorganic Chemistry, Vol. 11, No. 6,1972 1425

CATECHOL REACTIONS ion. A simple octahedral structure was suggested in which the catechol ions are bidentate ligands. Weiss, et a1.,2 suggested a binuclear structure but recently Flynn and Boer7 have determined the structure of the anhydrous pyridinium salt (CsHeN)z[Si(cat)a]by X-ray diffraction and shown that i t is mononuclear and has essentially the simple octahedral structure originally proposed by Rosenheim, et al. However, in a previous communication this authorg suggested structure I for

r

1

I

those complexes containing “water of crystallization,” including those prepared by Rosenheim and his coworkers. A similar structure, 11, has been proposed

H

0

4

I1

by Craddock and JoneslO for the arsenic(V)-catechol complex H [ A ~ ( c a t ).HzO. ~] Experimental Section Preparation of the Ethoxy Derivative. (NH&[Si(OCzHa)(Hcat)(cat)n].-A successful preparation requires rigorous exclusion of water. All operations were carried out in a dry nitrogen atmosphere using a vacuum line and the anaerobic techniques reviewed by Herzog, Dehnert, and Luhder.ll Ethanol was dried over calcium hydride and diethyl ether was dried over sodium. Both solvents were distilled as needed under dry nitrogen directly into the containers in which they were to be used. Glass containers were evacuated and heated with a heat gun before use in order to remove surface sorbed water. -4250-ml three neck flask containing 5.28 g of catechol (48.0 mmol) was fitted with a reflux condenser and a gas inlet tube with a stopcock. The flask was evacuated and then filled with nitrogen. Fifty ml of diethyl ether was distilled into the catechol. Another 50 ml of ether was distilled into an addition funnel and to this was added 2.5 ml of silicon tetrachloride (22.0 mmol). The addition funnel was fitted onto the third neck of the reaction flask and, while passing a slow stream of nitrogen through the system, the silicon tetrachloride solution was slowly added t o the catechol solution drop by drop. After addition was completed the solution was refluxed gently for 1 hr. The cooling water to the reflux condenser was turned off and the flow rate of nitrogen increased so that the ether distilled off. A viscous oil remained. The reflux condenser and addition funnel were replaced by stoppers and theremaining volatile material removed by vacuum. The vacuum must be applied slowly as serious foaming occurs. When foaming ceased and the pressure was less than 1 Torr the flask was heated on a steam bath for 30 rnin while pumping was con(9) D. W . Barnum, I n o v g . Chewt., 9, 1942 (1970). (10) J. H. Craddock and M. M . Jones, J. Arner. Chem. SOL.,83, 2839 (1961). (11) S. Herzog, J. Dehnert, and K. Liihder, “Technique of Inorganic Chemistry,” Vol. VII, H. B. Jonassen and A. Weissberger, Ed., Interscience, New York, N. Y., 1968, pp 119-149.

tinued. On cooling, the product was a hard resin with uncertain composition (H~[Si(cat)a]or ClSi(Hcat)s?). This resin was dissolved in 50 ml of dry ethanol and anhydrous ammonia passed in while stirring on a magnetic stirrer. A white precipitate of (NH4)z[Si(OC~H~)(Hcat)(cat)~] formed (structure 111)which was filtered off under nitrogen, washed with dry ethanol saturated with ammonia, and taken to constant weight under vacuum at room temperature. On contact with moist air it is rapidly converted to I. The X-ray diffraction patternI2 shows the three most intense lines a t d = 5.98 b (loo), 5.64 b (70), and 10.8 A (55) (numbers in parentheses are relative intensities), as well as a line a t d = 2.74 b that is due to ammonium chloride impurity. If the analytical data are corrected for the presence of ammonium chloride the formula is (NH4)~[Si(OC~H;)(Hcat)(cat)2] . Anal Calcd for (NH~)z[S~(OCZH~)(C~H~OZ)(C~H~O~)~] : Si, 6.46; C, 55.27; H, 6.03; N, 6.45. Found: Si, 6.32, 6.28; C, 52.66; H, 6.38; N, 6.99; C1, 1.92. Corrected for NH4Cl: Si, 6.49; C, 54.23; H, 6.34; N, 6.42. Preparationof ( N H l ) ~ [ S i ( c a t .)HzO. ~] (a) FromSilicic Acid.This preparation must be carried out in the absence of air since alkaline solutions of catechol are rapidly oxidized by oxygen. Enough silicic acid to provide 26.6 mg-atoms of silicon and 10.0 g of catechol (91.0 mmole) were weighed into a 250-ml flask with a side arm and a straight bore stopcock. The exact weight of silicic acid required was determined from its water content and a finely ground product was used in order to ensurerapid reaction. The flask was stoppered, evacuated, and filled with nitrogen, and 75 ml of 15 M ammonium hydroxide was added by means of a hypodermic syringe with the needle inserted through the bore of the stopcock. The contents were stirred in order to dissolve all of the catechol. Reaction appeared to be complete after 2-8 hr but the mixture was allowed to stand overnight in order to ensure complete reaction. The crystalline product was filtered off under nitrogen, washed with cold 15 M ammonium hydroxide, and dried under vacuum a t room temperature. Upon drying the crystals crumble to a white powder. The dry product was covered with ether and allowed to stand for 30 rnin in order t o extract small amounts of catechol impurity. It was then filtered again, washed with ether, and dried to constant weight under vacuum a t room temperature. The dry product can be handled in the open atmosphere without decomposition but it should be stored under nitrogen away from strong light. Because another complex containing 4 mol of catechol/g-atom of silicon sometimes coprecipitates, the X-ray diffraction pattern of each preparation should be examined for the absence of a line a t d = 9.1 A. The diffractipn pattern has two intense lines a t d = 11.46 b (100) and 5.77 A (95); all other lines have relative intensities less than 30. Analytical data are presented in Table I . If air has been carefully excluded during the preparation, I is a white powder. (b) From Colloidal Silica.-Two different sources of commercial silica sol were used: Ludox, containing 735 mg of SiOn/ml, from E . I. du Pont de Nemours and Co. and a sol containing 158 mg of SiOz/ml from Dr. Theo. Schuchardt Gmb.H., Munich, West Germany. The preparation used in this work was essentially the same as the preparation from silicic acid described above except that the reaction flask was first charged with 10.0 g of catechol (91.0 mmol), evacuated, and filled with nitrogen, and 100 ml of concentrated (15 M ) ammonium hydroxide was added by means of a hypodermic syringe with the needle inserted through the bore of the stopcock. The proper volume of colloidal silica containing 26.6 mg-atoms of silicon and diluted to 20 ml was added in the same way and the mixture set aside overnight to react. The crystalline product was recovered and dried to constant weight as described above. The crystals crumble to a white powder on drying. The product was frequently contaminated with the complex containing 4 mol of catechol/g-atom of silicon. The reaction was studied using different sources of silicon, different temperatures, various ratios of silicon to catechol, and various concentrations of ammonium hydroxide. As determined by X-ray diffraction patterns, numerous products were obtained, (12) The X-ray powder diffraction patterns and complete tabulations of the lattice spacings and relative intensities have been deposited as Document No. NAPS01648 with the ASIS National Auxiliary Publication Service, c / o CCM Information Corp., 909 3rd Ave., New York, N. Y . 10022. A copy may be secured by citing the document number and by remitting $2.00 for microfiche or 85.00 for photocopies. Advance payment is required. Make checks or money orders payable to: CCMIC-NAPS.

1426 Inorganic Chemistry, Vol. 11, No. 6, 1972 but when taken to constant weight the product was always I or a mixture of I and the catechol solvate. None of the intermediates was stable enough to isolate in a reliably pure form. At room temperature the reaction with either silicic acid or colloidal silica in 15 M ammonium hydroxide gives some visible crystals after 5-30 min and reaction appears to be complete after approximately 8 hr. Colloidal silica reacts somewhat faster than silicic acid. I n 10 M ammonium hydroxide several days is required for the reaction to go to completion. In 8 M ammonia the reaction is only partially complete after several weeks and in 6 M ammonia no crystalline product was obtained. T h e above procedure gave the same product if run at either 5 or 40'. About a 2070 excess of catechol is required to ensure that reaction with silica is complete. A larger excess of catechol is likely to cause formation of some of the catechol solvate (NH4)2[Si(cat)3]" ~ 0 'Hscat. The composition of the crystals changes rapidly upon removal from the mother liquor. Analysis shows that they are losing ammonia. For example in one preparation the product was washed with acetone and ether and then dried 2 hr in a slour stream of dry ammonia gas. The container was then swept with dry nitrogen for 20 min to remove the gaseous ammonia. This product contained 11.07, X. After drying 48 hr in a slow stream of dry nitrogen it contained 7.167, N, after 96 hr, 6.40y0 N, and after 144 hr, 6.367, N . In other cases the crystals were dried in a slow stream of nitrogen which was then bubbled through 47, boric acid solution and finally after the crystals were a t constant weight the ammonia was titrated with standard hydrochloric acid solution. Calculations of the amount of ammonia present in the original crystals gave values ranging from 2 to 4 mol of ammonia/g-atom of silicon. (c) From Silicon Tetrachloride.-The preparation given above for the ethoxy compound (XH1)~[Si(OC2Hj)(Hcat)(cat)~l was followed up to the formation of the resin. The resin was dissolved in absolute alcohol, and gaseous ammonia, saturated with water vapor by passing it through a tower of concentrated ammonium hydroxide, was introduced until no further precipitation occurred. Under these conditions the precipitate yields .HzO upon drying to constant weight. T h e (NHd)~[Si(cat)a] precipitate was filtered off under nitrogen, washed well with diethyl ether, and dried to constant weight under vacuum at room temperature. (d) Recrystallization of (NHd)~[Si(cat)~] .HzO .-This compound cannot be recrystallized from hot water because a precipitate of silica forms upon heating the aqueous solution. Recrystallization was achieved by dissolving 4.3 g of I in 40 ml of 6 ill ammonium hydroxide in a nitrogen atmosphere, filtering, and then passing in gaseous ammonia while cooling the solution to 0" in an ice bath. Colorless crystals of (NH,)t[Si(cat)sl. xHs0.yNH3 formed, which were filtered off, washed with cold concentrated ammonium hydroxide, and dried to constant weight under vacuum a t room temperature. Approximately 357, of the starting material was recovered. .HzO.-This compound is (e) Properties of (NHc)2[Si(cat)3] very soluble in water, dilute aqueous ammonia, ethanol and other alcohols, 2-ethoxyethanol, and dimethyl sulfoxide. It is insoluble in benzene, toluene, carbon tetrachloride, carbon disulfide, chloroform, dioxane, diethyl ether, acetonitrile, nitroethane, nitrobenzene, N,N-dimethylaniline, diethylamine, allyl chloride, and tert-butyl chloride. A small solubility was observed in tetrahydrothiophene 1,l-dioxide. I n aqueous acid it immediately gives a precipitate of hydrated silica. (NHr)2[Si(cat)a]1.5H20.-This compound was obtained on two occasions under quite different circumstances. 50attempt was made to find a reproducible synthesis. I n the first case the procedure described above for the preparation of I from silicic acid was followed except that 12 A l ammonium hydroxide was used for the reaction medium. The usual colorless crystalline product was obtained. After drying only 15 min under vacuum a t room temperature most of the material still had the shape of the original crystals; only a small amount hod disintegrated into a white powder. The product at this point gave an X-ray powder diffraction pattern with strongest lines a t d (h) = 10.8 (loo), 7.5 (30), 7.0 (25), 6.65 (60), and 6.41 ( 2 5 ) , and was stable on standing in air. However, upon taking this product t o constant weight under vacuum the diffraction pattern changed to that of I . In the second case, working under nitrogen, 2.0 g of I was dissolved in 22 ml of methanol (unknown water content), filtered, and set aside a t -5". Five days later colorless hexagonal

DENNISW. BARNUM crystals, 2-5 mm in diameter, were present. Upon drying 2 hr under vacuum at room temperature they disintegrated to a white powder and appeared to have reached constant weight. The X-ray diffraction pattern, however, showed it to be the sesquihydrate, A n d . Calcd for ( N H ~ ) Z [ S ~ ( C ~ H.~1 O . 5& H~ ] 0 : Si, 6.76; C, 52.03; H, 5.58; N, 6.74. Found: Si, 6.70; C, 32.37; H , 5.66; iY,6.73. Catechol of Crystalli~ation-(NH~)~[Si(cat)~] "10 .Hzcat. (a) From Silicic Acid.-The preparation was exactly the same as that of I described above, except that a large excess of catechol was used. Thus, 26.6 mg-atoms of silicon was taken, and 240 mmol (26.4 g) of catechol, which is 9 mol of catechol per g-atom of silicon. The reaction was carried out in 75 ml of 15 M ammonium hydroxide. A crystalline product was obtained in 1030 min, but generally the mixture was allowed to stand overnight. The crystals were filtered off under nitrogen, washed with cold concentrated ammonium hydroxide, dried 2 hr under vacuum a t room temperature, leached with ether, and dried to constant weight under vacuum a t room temperature. The crystals crumble to a white powder on drying. The X-ray powder diffraction pattern generally shows some I as impurity. (b) From Silicon Tetrachloride.-The preparation given above for the resin (HI [ S i ( ~ a t )or ~ ] CISi(Hcat)B?)was followed. This resin was dissolved in 50 ml of dry ethanol and divided into two equal portions. T o one portion gaseous ammonia saturated with water vapor was passed in which led to the preparation of I as described above. The second portion was held a t 0 ' in an ice bath while passing in gaseous ammonia saturated with water vapor. A fine white precipitate formed which was filtered off, washed with cold ethanol saturated with ammonia, and dried 1 hr under vacuum a t room temperature. The X-ray diffraction pattern of this product shows most intense lines a t d (b) = 10.5 (loo), 10.0 (BO), 9.2 (85), 5.8 (601, 4.35 (60),and 3.98 (40). I t was not identified further. Upon standing in the open atmosphere for 2 days the diffraction pattern changed to that of (NH4)2[Si(cat)8].H20 Hncat. T h e product can be handled in the atmosphere once it is dry; however it discolors in a few days. It is best stored under nitrogen in the dark. The X-ray powder diffraction pattern has a single intense line a t d = 9.1 -&. All other diffraction lines have relative intensities 25 or less. 4 n a l . Calcd for ( J S H ~ ~ Z [ S ~ ( C ~ H "zO.CBHBOZ: ,OZ)~] Si, 5.44; C, 55.80; H , 5.46; N, 5.42. Found for three different preparations: Si, 5.49, 5.52, 5.41; C, 54.96, 54.84, 55.47; H , 5.58, 5.47,5.76; X', 5.73,5.84,6.39. Preparation of the Dimethyl Sulfoxide Solvate (NHc)2[Siat)^] ~H20.2CH3SOCHI.-Thefollowing operations werecarried out in the open atmosphere without noticeable decomposition. Dry reagent grade dimethyl sulfoxide (10 ml) was used to dissolve 1.0 g of I, and the resulting solution was filtered through highly retentive filter paper. While stirring, 25 ml of chloroform was added drop by drop. A precipitate formed after addition of about 10 ml of chloroform. It was filtered off, washed with chloroform, and dried under vacuum at room temperature to constant weight. An ethanol solution of this compound was examined by tic on alumina eluted with 90yc toluene-lO~,, ethanol. Two spots were observed; one with the same RF as dimethyl sulfoxide, and one with the same l i p as I. Anal. Calcd for ( N H ~ ) z [ S ~ ( C B H .~HOzZO).~ ]( C H & O C H B ) : Si, 4.99; C, 46.95; H , 6.09; X, 4.98; S, 11.39. Found for four separate preparations: Si, 5.49, 5.11, . . . , 45.35, 48.92, 46.95; H, 6.25, 6.22, 5.93, 5.85; S , 4.69, 4.75, 5.23, 5.15; S, 10.54, 10.79,11.14,11.39. The X-ray powder diffraction pattern has several intense lines at d (A) = 11.0 (95), 10.1 (35), 7.6 (40), 6.7 (351, 5.12 (301, 4.98 (loo), 4.57 (SO), 4.50 (30), 3.61 (45), and 3.50 (40). T h e infrared absorption spectrum is very similar to that of I except for bands a t 9.9 and 10.5 p that are due to the dimethyl sulfoxide. Crystals of a size suitable for X-ray diffraction studies can be grown by placing a solution of I in dimethyl sulfoxide inside a desiccator containing chloroform in the bottom. Isothermal distillation of chloroform into the dimethyl sulfoxide gradually occurs and crystals begin to form in about 10 days. Preparation of the Ethanol Solvate ( N H 4 j ~ [ S i ( ~ a t x) 3H] ~ O C2HjOH.-This product is obtained whenever (NHi)z[Si(cat)31. H20 crystallizes from an aqueous ethanol solution. The procedure given above for the preparation of (XHa)z[Si(cat)8] .H20 from silicic acid was followed except that a mixture of 25 ml of

Inorganic Chemistry, Vol. 11, No. 6, 1972 1427

CATECHOL REACTIONS

TABLE I ANALYTICAL DATAFOR (NH& [Si(cat)a] H 2 0 e

Theory

r--Prepn Ref 3

from SiC14-This work

Prepn from Si08 sol

Si C H N 0“

a

Prepn from silicic d a-ci This work-

,

6.91 7.04 6.86 6.81 7.01 53.18 52.83 52.39 53.04 5.46 5.40 5.56 5.49 6.89 6.73 6.48 6.58 27.55 (28.2) (28,s) (27.9) “4 8.88 8.88 Values for oxygen in parentheses are calculated by difference from 100%.

ethanol and 75 ml of 15 M ammonium hydroxide was used for the solvent. Three to five days a t room temperature was required for the reaction t o go to completion. The crystalline product was filtered off under nitrogen, washed with cold concentrated ammonium hydroxide, dried 5-10 min under vacuum, and finally dried to constant weight by standing over calcium sulfate in a desiccator filled with nitrogen. When dry it can be handled in the air without decomposition. The nmr spectrum shows one ethyl group per three molecules of catechol with the methyl protons centered at r 9.00 ppm and the methylene protons centered a t T 6.60 ppm relative to TMS. The X-ray powder diffraction pattern has most intense lines a t d (A) = 10.0 (loo), 8.75 (30), 8.2 (20), 5.67 (25), 5.09 (25), 3.83 (50), and3.15 (25). The analytical data are not good enough to distinguish between 3 and 4 mol of water of crystallization. The best fit however is obtained assuming four “water” molecules. Anal. Calcd for (NHa)z[Si(cat)s].4H?O,CzHeOH: Si, 5.75; C, 49.16; H, 6.60; N, 5.73. Found: Si, 5.65, 5.77; C,47.20; H,6.59; N,5.73. On standing under vacuum this compound loses both water and ethanol, the final product being the “monohydrate,” I . Large crystals (2.5 mm long) can stand in the open air for many days without noticeable decomposition and after 1 year in a closed vial were only partially converted to I. However, if ground to a fine powder it is completely converted t o I by standing in air 2 days or by several hours under vacuum. The ethanol solvate was also prepared on several occasions by adding ethanol to the filtrate obtained in the preparation of I from silicic acid or colloidal silica in concentrated ammonium hydroxide solution. About 50 ml of ethanol was added to the filtrate and the solution was stored overnight a t 5’. Also, the ethanol solvate can be prepared by dissolving 1.0 g of I in 30 ml of 50:50 ethanol-water (by volume) containing 1 ml of 15 M ammonium hydroxide and allowing this solution to stand overnight or longer a t - 5 ” . Preparation of the Acetone Solvate (NHa)~[Si(cat)~] “20. CH&OCH3.-In some of the first preparations done in this work the crystalline products obtained in the reaction of silicic acid or colloidal silica with aqueous catechol were washed with acetone and then with ether. I t was observed that more crystals form in the filtrate upon addition of the acetone. In three cases these crystals were also recovered, washed with acetone, and dried 1 hr under vacuum a t room temperature. The X-ray powder diffraction pattern of the dried material has a single intense line a t d = 10.7 A (100). Of the remaining lines the more intense ones occur a t d (A) = 7.9 (12), 6.45 (14), 4.95 ( 2 5 ) ,and 3.35 (10). The ir spectrum shows only a weak band a t 5.9 p due to the C=O stretching band of acetone. The remainder of the ir spectrum is essentially the same as that of I . A n d . Calcd for (NH&[Si(cat)s] .H%O.CH&OCHa: Si, 6.05; C, 54.29; H, 6.07; N, 6.03. Found for three different preparations: Si, 6.11, 5.99, 5.78; C, 52.74, 53.44, 54.56; H, 6.28, 6.02, 5.96; N,6.27, 6.33, 5.86. This compound was also obtained simply by covering I with acetone and allowing the mixture to stand for a few minutes. There is no visible change, but after such treatment the diffraction pattern is that of the acetone solvate. Acetone can be removed from the solvate simply by washing with ether. Obviously, acetone should not be used as a wash liquid; some of the diffraction patterns of early preparations of I in which this was done show the presence of acetone solvate as impurity.

Structure of ( N H & [ S i ( ~ a t ).H,O ~l In this work all of the crystalline products obtained from the reaction of catechol with silica in aqueous ammonium hydroxide contained ammonia of crystalliza-

6.93 52.95 5.32 6.15 (28.6)

6.98 52.01 5.49 6.67 (28.8)

-6.92

Ref6

-

7.10 54.69 5.40 6.93 (25.88)

tion and perhaps more than 1 mol of water of crystallization. They begin to lose ammonia as soon as they are removed from the mother liquor so i t was not possible to establish an exact composition. The general formula, however, is (“4)~ [ S i ( ~ a t ).xHzO ~ ] .yNH3, where x = 1 or 2 and y = 2-4. Other catechol complexes containing ammonia of crystallization have been reported previously, namely, RZ[ T i ( ~ a t ) 5NH3l3 ~] and H [ A ~ ( c a t )H ~ ]z O . ~ / ~ N HIt~ is . ~likely ~ that different amounts of water and ammonia are obtained depending on the conditions of the reaction. Many different X ray powder diffraction patterns were obtained but no reproducible behavior was evident. The diffraction patterns are complex, containing many lines, and appear to depend on the concentration of ammonia in the solution from which the crystals formed as well as the drying time. Nevertheless, all of these products gave the diffraction pattern of (NH4)~[Si(cat)s]. HzO after taking to constant weight (although sometimes the catechol solvate (“4)Z [ S i ( c a t ) ~ ] . H ~ O . H ~was c a t present as an impurity) .le Almost identical behavior has been observed in the reaction of catechol with niobium and tantalum hydroxides. Analytical data for several preparations of (NH4)z[Si(cat)r]~ H z Ofrom silicic acid, colloidal silica, and silicon tetrachloride are given in Table I. The data agree best with a complex containing 1 mol of water of hydration. Previous data on the amount of water are conflicting. Rosenheim and Sorge3 found 9 mol of water and Weiss, et al.,, found 4.5 but neither of these groups dried their product to constant weight nor did they report ammonia of crystallization in their products. Hess, et a1.,6found 1 or 2 mol of water. Weiss, et al., found that the silicon and germanium complexes are isomorphous. Unfortunately they gave no analytical data for their germanium complex, but Bevillardl7 found (NH4)~[Ge(cat)3]* 1.5H.20. A silicon complex with 1.5 mol of water was also isolated in this work and, as shown by its X-ray diffraction pattern, is definitely different from the monohydrate. Rosenheim and Sorge3claimed to have prepared anhydrous (NH&[Si(cat)3] from the reaction of catechol with silicon tetrachloride. Since their preparation was supposedly done in the absence of water, i t is not surprising that they did not consider the possibility of a hydrate. However, in this work i t was found that one must exclude water rigorously if the monohydrate is to be avoided, and even then the product is (“&[Si(13) c.Gopinathan and J. Gupta, Indian J. Chem., 3,470 (1965). (14) R . F.Weinland and J. Heinzler, Bey., 82, 1316 (1919). (15) The ir spectra cannot be used t o distinguish the various products as the differences are too slight and occur mainly in the N-H and 0-H stretching region where there is a great deal of overlapping. (16) F. Fairbrother, N. Ahmed, K. Edgar, and A. Thompson, J. Less-

Common Metals, 4,466 (1962). (17) P. Bevillard, Bull. Soc. Chim. Fv., 296 (1954).

1428 Inorganic Chemistry, Vol. 11, No. 6, 1972

(OCZH~) (Hcat) cat)^] (structure 111). Upon contact

1

I11

with moist air the ethoxy compound is converted within 30 min into the monohydrate, I. Since Rosenheim and Sorge did not mention any special precautions to avoid contact with water and since their analytical data are in better agreement with a monohydrate than an anhydrous complex (see Table I), i t is probable that they had, in fact, prepared the monohydrate. Flynn and Boer’ have established the structure of the anhydrous pyridinium salt (Cd%N)2[Si(cat)s] by an X-ray diffraction study of a single crystal, and in this laboratory i t has been found that the water can be removed from Mg[Si(cat)s].HzO by heating a t 100”under vacuum for 166 hr. In a previous communicationg structure I was proposed for the “monohydrate.” One would tend to question this formulation in view of the X-ray structure determination on the pyridinium salt ; however, the evidence for structure I is strong and i t still deserves serious consideration for those compounds containing “water of crystallization.” The main evidence for structure I was obtained from the reaction of silicon tetrachloride with catechol in anhydrous ether. The hydrogen chloride that was produced by the reaction, as well as the ether, was removed by sweeping with nitrogen and finally by vacuum. A resinous product remained which had an uncertain composition. Weiss, et ~ l .and , ~ Schwarz and Kuchenlx claimed i t to be H ~ [ S i ( c a t )but ~ ] Rosenheim and Sorge3claimed i t to be ClSi(Hcat)a. None of these workers presented convincing data and work in this laboratory so far has given conflicting results. If gaseous ammonia is passed into a solution of this resin in anhydrous ethanol, an ammonium salt precipitates which contains an ethyl group and probably has structure 111. It is especially sensitive to moisture and is completely converted to I on contact with the open atmosphere for 30 min. The water in (NI-11)2[Si(cat)3].H20 is much more strongly held than one would expect for simple water of crystallization. Thus, heating under vacuum ( cis complexes were obtained. When the L the product was established by its melting point (211 f 2', lit.a is P(n-C4H9)3, a trans complex is obtained. We sugmp 211-213') and by a comparison of the 'H nmr with the reported values. The 'H nmr spectrum (CSz) exhibited absorptions (1) M. F. Hawthorne, D. C. Young, T. D. Andrews, D. V. Howe, R. L. a t 6 3.50 (broad singlet, intensity 1, due to the C-H on methylPilling, A . D. Pitts, M . Reintjes, L. F. Warren, Jr., and P. A. Wegner, J . carborane(l2)), and 6 2.00 (broad singlet, intensity 3, due to the Amer. Chem. Soc., 90, 879 (1968); M. F. Hawthorne and T. A. George, i b i d . , C-methyl protons) (litas6 3.48 and 1.48, respectively). The pure 89, 7114 (1967); T . A. George and M. F. Hawthorne, ibid.,90, 1661 (1968); rnethylcarborane(l2) was stored in a n evacuated desiccator over G . Popp and M. F. Hawthorne, ibid., 90, 6553 (1968); M. F. Hawthorne and PZOSprior to use. H. R. Ruhle, Irzoug. Chem., 8 , 176 (1969); T. A. George a n d M . F. Hawthorne, J . Amer. Chem. SOL.,91, 5475 (1969); M. F. Hawthorne, Accounts Synthesis of the Cis Platinum-Phosphine Complexes.-The Chem. Res., 1, 281 (1968). following platinum compounds cis-dichlorobis(trimethy1phos(2) S. Bresadola, P. Rigo, and A. Turco, Chem. Commun., 20, 1205 (1968); phine)platinum(II) [czs-PtCl~[P(CHa)3]~] , cis-dichlorobis[diTHE

J. C. Smart, P. M. Garrett, and M. F. Hawthorne, J . Amev. Chem. Soc., 91, 1031 (1969); C. M . Mitchell and F. G. A. Stone, Chem. C o m m u n . , 22, 1263 (1970); D. A . Owen a n d M. F. Hawthorne, J . Amev. Chem. Soc., 92, 3194 (1970).

(3) M. F. Hawthorne, T. D. Andrews, P. M. Garrett, F. P. Olsen, M. Reintjes, F. N. Tebbe, L. F. Warren, P. A . Wegner, a n d D. C. Young, Inorg. Syn., 10, 9 1 (1967).